ComfyUI_MCP
3
ComfyUI_MCP Server is a Model Context Protocol server for ComfyUI, designed to be extensible and configuration-driven. It enables automation and integration of AI tools like txt2img and img2img, supporting custom workflows through JSON configurations and decorator modules.
Overview
ComfyUI_MCP Server is an extensible and configuration-driven Model Context Protocol server designed for ComfyUI. It supports MCP services like txt2img and img2img, which can be customized through JSON and extended with MCP tool decorators. This makes it suitable for automation and integration in AI applications like image generation. Key features include tool auto-registration, configuration-driven parameters, and integration with MCP Inspector for debugging.
Installation & Configuration
- Run
python install.py
for installation and configuration. - Use
run_mcp.bat
on Windows or./run_mcp.sh
on Linux to start the server.
Tool Auto-Registration & API Extension
- Add custom MCP tools by implementing specific Python files and JSON configurations.
- Automatic effect upon addition without modifying the main entry.
Resource Registration & Management
- Register resource-type APIs for automatic management using decorators.
Debug with MCP Inspector
- Use MCP Inspector for interactive tool testing and debugging.